very low efli

[johnmf]

hi......

ive got a 3 phase db to do some work on at work.

the mains supply is 3 phase (185mm conductors) protected by 315A bs88's which goes through a mains isolator. this isolator feeds a busbar chamber off which the board that needs working on is running.

problem is that the earth fault loop impedance readings at the board were below 0.01!!!!

likewise the pfc ( between any phase and earth) was above 19KA!!

at first these readings seemed unlikely but then i remembered that the substation is actually in the premises of the factory(about 20 metres away)!

wanted to fit a new board but dont think they'l be rated to 19KA. any ideas anyone?

 

[Pensdown]

If your calculations are correct it looks like you'll be using MCCB's with a 25kA rating

 

[johnmf]

where's the best place to look for these if you dont mind me asking.

what makes am i to look out for/avoid?

if the fault current to earth is above 19KA then the recorded fault currents going to have to be at least double that (phase to phase)

am i correct in thinking this?

funny thing is we had a PIR done year ago and the guy has the ze as 0.07.
im thinking he must have made this up because there's no way it can be that given the short length to the substation and the cable size!

is my only option a fuse board that takes bs88's?

 

[Spark123]

A BS88 fuseboard would be imo the way to go, GE redspot seem to be a reputable brand. What size circuits are you running from the board?

 

[Adam_151]

Do a low ohms test on the submain cable (both R1 + R2 and R1 + Rn) and see what that comes out as, OOI , I'm a tad suspicious of ELFI testers at the best of times... (A usful tool no doubt, but like only instrument... not beyond questioning)

 

[Pensdown]

If your calculations are correct it looks like you'll be using MCCB's with a 25kA rating

I should have added....or good old BS88 fuses.

I agree with Adam and I would double check even though the results may still be very low

 

[johnmf]

know what your saying guys about the efli testers. had that thought myself but what with the location of the supply and cable sizing im fearing the worst.

circuits from the db are some small extrxactor motors(3 phase), some welding socket outlets and some ring circuits.

i wanted to rcd the ring circuits because they arent currently. this could now be a problem!

cant understand though how the PIR man got the results he did, ive checked his efli results on lots of circits and they are all lots lower than he's stated.

 

[RF Lighting]

was his / has your tester been recently calibrated?

 

[Adam_151]

What sise is the submain, and what the the cpc? just the armour or a core as well (assummig SWA?) and how long is it? a quick calculation for that part of the loop will give you an estimate that should be pretty close to the mark for that (which can be backed up by measurement of the same, which is going to be close to the mark) if you get a value higher than from the loop test, then assume worst case ze (0) [though if you know some details about he transformer, a calculation would probably be possible] and do Ze + (R1 + R2)

If you really have such a big potential fault current then I'd choose fuses over breakers, you'll most likely find that they will do a much better job at keeping the let-through energy down than high Ka breakers (the reason is simple, there is a limit to how fast breaker contacts can swing, with a fuse, the let through (I²t) is constant becuase you need to disapate a certain amount of energy in a piece of wire to vapourise it once you start doing it do fast that the heat can't really excape at all)

 

[pdcelec]

Is your tester a Fluke? Can expand on this question if required.

 

[johnmf]

the meter is megger mft1502, and it was calibrated 3 months ago. brand new infact.

thanks for all the help so far guys, its all info thats greatly appreciated as ive never come across this problem before to be honest!

 

[pdcelec]

have you removed paralell earths?

 

[johnmf]

to be honest no. got no chance with the size of the install and the downtime demand!

 

[pdcelec]

Have you measured the R1+R2 values?

 

[dunga7]

hi just a quick note, with the megger 1502 I think it is possible to measure your phase to phase pfc instead of doubling your highest phase/ earth.